--- title: "How Birth Time Rectification Works: Evidence, Not a Verdict" slug: "how-birth-time-rectification-works" canonical_url: "https://mahadasha.com/how-birth-time-rectification-works" description: "How birth time rectification works: candidates ranked by evidence, never a confidence percentage. A D60 lagna holds for 1.33 minutes, and that sets the limit." published_date: "2026-08-06 00:00:00" modified_date: "2026-08-06 00:00:00" author: "Pawan Acharya" keywords: "how birth time rectification works, birth time rectification, unknown birth time astrology, rectify birth time, d60 lagna, kp sub lord rectification" locale: "en" --- # How Birth Time Rectification Works: Evidence, Not a Verdict > Rectification narrows a remembered birth time by asking which candidate produces a chart that agrees with things already known. What it can settle, and where it stops. # How Birth Time Rectification Works: Evidence, Not a Verdict How birth time rectification works, in one sentence: you take a window of candidate instants around the time you were told, and you ask which of them produce a chart that agrees with things already known about the person. It is a craft rather than a solved problem, and the single most important design decision in a tool that does it is what the tool refuses to output. The refusal here is a confidence percentage. There is no honest way to put a probability on a birth time, and the engine that came before this one produced its confidence figures with a random number generator, capped at 95 so they always read as credible. That is worth knowing about any tool that offers you a number: ask what produced it. ## Why ten minutes matters more than you would think The case for rectification is a table, not an argument. Take a chart and shift the birth time by a fixed amount, then count what changes: - 1 minute: rising sign changes 0.80 percent of the time, navamsa lagna 7.18 percent, KP sub-lord 17.50 percent - 4 minutes: 3.32, 29.65 and 61.63 percent - 10 minutes: 8.80, 75.30 and 99.78 percent - 60 minutes: 50.88 percent, and the tenth divisional chart is now wrong 91.02 percent of the time against a pure-chance figure of 91.67 So a rough birth time is not uniformly rough. The rising sign is robust, the [navamsa](https://mahadasha.com/navamsa-chart-d9) is fragile, and the KP sub-lord is gone almost immediately. Anyone reading a divisional chart off a time rounded to the nearest quarter hour is reading a chart that belongs to somebody else three quarters of the time. ## How birth time rectification works, in three steps ### Step one: candidates, not a search A run starts with an anchor time and a window, say two hours either side, and a step. Candidates are generated as signed offsets from the anchor and converted to wall clock at the very end, which sounds fussy until you see what the alternative did. Walking the clock forward from 23:30 wrapped past midnight and never terminated, and the same code given 01:00 with a two-hour window produced exactly one candidate, 23:00, which it then reported as the rectified birth time with a confidence figure attached. Each candidate carries its own calendar date, because a candidate two hours after 23:30 belongs to the next day, and charting it on the anchor's date puts the ascendant most of a day out of place. ### Step two: find the edges exactly The naive approach recomputes the chart at every second and looks for changes. For a two-hour window that is 7,200 charts, it is far too slow to feel interactive, and it still only locates a boundary to the nearest second it happened to sample. There is a better route. The ascendant increases strictly with time, so a division boundary is a root rather than a target to hunt for. Bisect it and any crossing instant falls out in about fifteen ephemeris calls, exactly rather than approximately. That single routine answers everything the workspace needs: which cell am I in, where does it start, where does it end, and how far am I from the edge. Knowing the distance to the edge is the part people underuse. A candidate sitting four seconds from a navamsa boundary is a different proposition from one sitting in the middle of its cell, and no ranking that only reports the winning sign will tell you which you have. ### Step three: what counts as evidence Three kinds of evidence are available, and they are weighted differently for a reason. Divisional charts come first, weighted by Vimshopaka Bala, the classical Shodashavarga scheme that totals 20. That weighting is citable rather than invented, and it produces some counterintuitive results: the [D60](https://mahadasha.com/shashtiamsa-chart-d60) outranks the career chart by eight to one, which is the opposite of how the two are usually treated in practice. Five charts outside Parashara's sixteen carry no classical weight at all, so when somebody selects them they get a flat token value that is labelled as this application's convention rather than dressed up as tradition. The KP sub-lord of the ascendant comes second, because it is the finest commonly used division and therefore the most discriminating. Life events come third, and they are the weakest link in every rectification system ever built, including this one. Matching a remembered year to a period boundary is easy to do and easy to do badly, since the periods are wide enough that something plausible is always running. The method weights that combine these three are this application's, not anybody's tradition, so they are returned in every response along with the per-method scores. A reader who disagrees with the weighting can recompute the ranking themselves, which is the only defensible way to ship a number nobody can prove. ## What the answer is allowed to say The output is a ranked list of candidates with the evidence behind each, plus a resolution statement naming the finest chart tested and the span it can actually distinguish. That last part is the piece most tools omit and the piece that makes the rest honest. Measured at Delhi over a full day, a rising sign holds for between 84.92 and 140.33 minutes. A navamsa lagna holds for 9.25 to 15.67 minutes. A KP sub cell holds for 1.50 to 10.42 minutes. The sixtieth-part lagna, the finest division offered, holds for 1.33 to 2.42 minutes. So a result of 14:32:07 does not mean 14:32:07. It means somewhere inside a cell roughly two minutes wide, and presenting the seconds without the cell is a precision claim the method cannot support. Nothing in the response converts the ranking into a percentage, and the resolution statement rather than the score is what tells you how much precision you have earned. ## What it cannot do Rectification narrows. It does not discover. Given a window of two hours it can often reduce the plausible range to a handful of cells, and given no anchor at all it can do very little, because the space of candidates becomes a whole day and the evidence does not scale with it. It is also only as good as the events you feed it, and human memory for dates is worse than people believe. A marriage year is usually solid. A month when a job changed is often out by one. If your evidence is soft, the honest use of the tool is to see which candidates survive rather than to pick the top row. When you have no usable birth time at all, the alternative is not a better rectification, it is a different branch of the subject: [prashna](https://mahadasha.com/what-is-prashna-kundali) answers a single question from the moment it is asked and never touches the birth chart. ## Running one The [rectification tool](https://mahadasha.com/birth-time-rectification) takes your anchor time, a window, and any life events you are confident about, and returns the ranked candidates with the reasons attached. The workspace beside it lets you step the time by a second at a time and watch which divisions change, which is the part that teaches the most. If you have not yet cast the chart at all, the [Birth Chart Calculator](https://mahadasha.com/birth-chart-calculator) is the place to start. ## Frequently asked questions about rectification ### Can you explain how birth time rectification works in one paragraph? Generate candidate times around the one you were given, cast a chart for each, and keep the ones whose divisional lagnas and sub-lord agree with what is already known. Rank by weighted evidence, then state how fine a distinction the charts you used can actually make. With no anchor at all the candidate space is the whole day and the method has very little to work with. ### Why is there no confidence percentage? Because no honest one exists, and because the previous engine generated its percentage at random and capped it at 95 so it always looked plausible. What replaces it is the evidence per candidate and a statement of the finest span the method can distinguish. ### Which divisional charts should I select? Start with the navamsa and the tenth, because their cells are wide enough that a match means something. Adding the sixtieth only helps if your anchor is already good to a couple of minutes, since its cells are barely wider than that. ### Can it use my life events? Yes, and treat that part with the most suspicion. Period matching is the weakest of the three evidence types because the periods are wide and something is always running. It is a tiebreaker, not a foundation.